***** Codebook for “Do Elite Partisan Disagreements Degrade the Probability of Victory in War? Evidence from South Korean Military Battle Experiments” by Michael F. Joseph, Joon H. Chung, and Hui Seong Park


*** milexp1: Dataset for experiment 1. Experiment 1 was conducted in person at the shooting range before junior and senior cadets participated in the shooting drill. Only the Korean Army cadets took experiment 1.

** Behavioral measures

* order-3: Serial numbers assigned to record the order of survey questionnaires turned in for junior cadets.

* order-4: Serial numbers assigned to record the order of survey questionnaires turned in for senior cadets.

* shooting: Shooting score (maximum 20; feasible value ranges from 0 to 20, depending on how many out of 20 targets a respondent shot). Only senior cadets participated in the battlefield shooting drill, while juniors received practice shooting training.

** Treatment assignment

* group: Treatment assignment identifier, depending on the values of the two independent variables (elite partisan disagreement and respondent’s party affiliation) assigned in the provided scenario. 1 if partisan agreement and affiliated with the incumbent party, 2 if partisan agreement and affiliated with the opposition party, 3 if partisan disagreement and affiliated with the incumbent party, and 4 if partisan disagreement and affiliated with the opposition party.

** Mechanisms

* NIM: Perception of national interest, measured by asking to what extent a respondent agrees/disagrees that the war depicted in the treatment scenario serves South Korea’s national interest. Likert scale ranging from 1 (strongly disagree) to 7 (strongly agree).

* TIM: Soldier-to-soldier trust, measured by asking to what extent a respondent agrees/disagrees that they trust other soldiers across the Korean military would fight to their full potential when the war in the treatment scenario happens. Likert scale ranging from 1 (strongly disagree) to 7 (strongly agree).

** Dependent variable

* SLE1-risklife: Military cohesion (main dependent variable) indicator. The extent to which a respondent is willing/reluctant to risk their own life by volunteering to serve in a risky military mission. Likert scale ranging from 1 (strongly reluctant to) to 7 (strongly willing to).

* SLE2-risksubordin: Military cohesion indicator. The extent to which a respondent is willing/reluctant to risk their subordinates’ lives by ordering them to serve in a risky military mission. Likert scale ranging from 1 (strongly reluctant to) to 7 (strongly willing to).

* SLE3-discretion: Military cohesion indicator. The extent to which a respondent is willing/reluctant to use their discretion to accomplish the operational objectives. Likert scale ranging from 1 (strongly reluctant to) to 7 (strongly willing to).

* SLE4-interunit: Military cohesion indicator. The extent to which a respondent trusts/mistrusts other units and is willing/reluctant to cooperate with them to accomplish the operational objectives by doing what they promised to do (advancing at 11pm in the given scenario). Likert scale ranging from 1 (I would almost definitely send a lookout and advance if others are.) to 7 (I would almost definitely advance at 11pm.).

* SLE5-unitcohesion: Military cohesion indicator. Unit cohesion, measured by the likelihood of volunteering for a high-risk mission when others in their service academy cohort asks them to join the mission and fight together. Likert scale from 1 (very unlikely) to 7 (very likely).

* SLE6-obey: Military cohesion indicator. The willingness to obey the superior’s command even when one thinks it is unreasonable. Measured as the perceived appropriateness of the hypothetical command to fight a battle when a respondent’s platoon has already fought five high-risk missions. Likert scale from 1 (very inappropriate) to 7 (very appropriate).

** Alternative mechanisms and additional checks on mechanisms

* party1: The extent to which a respondent agrees/disagrees that, in a war like what was described in the treatment scenario, they would like to fight alongside soldiers who share their political beliefs. Likert scale ranging from 1 (strongly disagree) to 7 (strongly agree).

* party2: The extent to which a respondent agrees/disagrees that, in a war like what was described in the treatment scenario, they would avoid fighting alongside soldiers with the opposite political beliefs. Likert scale ranging from 1 (strongly disagree) to 7 (strongly agree).

* party3: The extent to which a respondent agrees/disagrees that, in a war like what was described in the treatment scenario, they would like to fight alongside soldiers from the region they themselves are from. Likert scale ranging from 1 (strongly disagree) to 7 (strongly agree).

* opposition1-genuine: The extent to which a respondent thinks that the opposition party’s stance in the treatment scenario they are exposed to reflects their genuine concerns about national interest. Likert scale from 1 (none at all) to 5 (a great deal).

* opposition2-partisan: The extent to which a respondent thinks that the opposition party’s stance in the treatment scenario they are exposed to reflects their domestic political concerns regarding partisan struggle. Likert scale from 1 (none at all) to 5 (a great deal).

* AM1-threat: The extent to which a respondent agrees/disagrees that China poses a security threat to Korea. Likert scale ranging from 1 (strongly disagree) to 7 (strongly agree).

* AM2-partytype: The extent to which a respondent agrees/disagrees that the party they favor can plausibly take the stance stated in the treatment scenario. Likert scale ranging from 1 (strongly disagree) to 7 (strongly agree).

* AM3-artificial: The extent to which a respondent agrees/disagrees that the treatment scenario was plausible and not too unrealistic. Likert scale ranging from 1 (strongly disagree) to 7 (strongly agree).

** Demographics

* age: A respondent’s age in year, feasible range set to be from 18 to 60.

* female: 0 if a respondent’s biological sex given by birth is male and 1 if female.

* rank: Identifier of a respondent’s rank in terms of military hierarchy. 1 if junior cadet and 2 if senior cadet.

* volunteer: 1 if a respondent volunteered to serve in the military and 0 otherwise.

* edu: The level of educational attainment. 1 if graduated high school, 2 if holds an associate’s degree, 3 if holds a bachelor’s degree, and 4 if holds an advanced (graduate) degree.

* hometown: Identifier for the province in which a respondent’s hometown is located. We intend to capture the regional divide between Gyeongsang province and Jeolla province, which are known for the long history of hostility against each other. 3 if a respondent is from Jeolla province, 2 if from Gyeongsang province, and 1 if from the rest (including foreign countries).

* sibling: The number of siblings a respondent has. 0 if a respondent is the only child, 1 if has one sibling, 2 if has two siblings, and 3 if has three or more siblings.

* partyIDstrength: The degree of self-identification with the party a respondent favors. Likert scale ranging from 1 (Very weakly identify with the party; I would rather identify myself as an independent) to 7(Very strongly identify with the party).

* priorexp: 0 if a respondent has no military experience before attending the service academy and 1 if they do (e.g. as a conscripted rank-and-file soldier or a non-commissioned officer).

* cat: 0 if a respondent has no experience of having a cat as a pet while growing up and 1 if they do.

** Psychological attributes

* anger: The frequency of getting angry at people and situations. Likert scale from 1 (never) to 7 (every time).

* risk: Risk propensity measured by how easy or difficult it is for a respondent to accept taking risk. Likert scale from 1 (very easy) to 7 (very difficult).

* moral: Moral consciousness measured as the frequency of making choices that go against personal interest to do what seems right. Likert scale from 1 (never) to 7 (every time).

* dove: Dovishness measured by the extent to which a respondent agrees/disagrees that the use of force only makes problems worse. Likert scale from 1 (strongly disagree) to 7 (strongly agree).

-------

*** milexp2: Dataset for experiment 2; here we only list variables that are newly added in the experiment 2; all other variables in the data set are the same that appear in the milexp1 data set. Cadets, conscripted rank-and-file soldiers, and active duty commissioned officers, all affiliated with the Korean Army, participated in experiment 2.

* Duration: The amount of time spent for completing the survey (in seconds). Experiment 2 was conducted online via Qualtrics, so the duration data is available.

* open1: Open-ended survey response for the question, “Imagine the possible situation in which the president orders you to fight in a war, but the opposition party disagrees. In this case, would you have any concern about how other soldiers would fight? If so, what are those concerns?”

* open2: Open-ended survey response for the question, “Do you think it is important that bipartisan agreement is reached about the decision to fight a war? If so, what concerns would you have about fighting a war if one party supports it but the other does not?”

* open3: Open-ended survey response for the question, “In your opinion, what factors are most effective in motivating you to fight in earnest in war?”

* rank: A respondent’s military rank. 1 if conscripted rank-and-file, 2 if junior cadet, 3 if senior cadet, 4 if second lieutenant, 5 if first lieutenant, 6 if captain, 7 if major, 8 if lieutenant colonel, and 9 if colonel.

* class: Identifier for the class of respondents. “cadet” if a respondent is a junior or senior cadet, “soldier” if a conscripted rank-and-file soldier, “officer1” if an active duty commissioned officer stationed in the Korean Army Academy, and “officer2” if an active duty commissioned officer stationed in different Army bases.

-------

*** pilot_data: First round of pilot survey with Korean Army cadets. Used for checking whether and to what extent there exists variation in responses to scenario-based questions measuring military cohesion (DV). Cadets were randomly assigned to two among six scenario-based DV measuring questions (first set, second set).

** First set of DV measuring tasks

* riskownlife: The extent to which a respondent is willing/reluctant to risk their own life by volunteering to serve in a risky military mission. Standardized Likert scale ranging from -3 (strongly reluctant to) to 3 (strongly willing to).

* riskotherlife: The extent to which a respondent is willing/reluctant to risk their subordinates’ lives by ordering them to serve in a risky military mission. Standardized Likert scale ranging from -3 (strongly reluctant to) to 3 (strongly willing to).

* discretion: The extent to which a respondent is willing/reluctant to use their discretion to accomplish the operational objectives. Standardized Likert scale ranging from -3 (strongly reluctant to) to 3 (strongly willing to).

* coordination: The extent to which a respondent trusts/mistrusts other units and is willing/reluctant to cooperate with them to accomplish the operational objectives by doing what they promised to do (advancing at 11pm in the given scenario). Standardized Likert scale ranging from -3 (I would almost definitely send a lookout and advance if others are.) to 3 (I would almost definitely advance at 11pm.).

* unitcohesion: Unit cohesion, measured by the likelihood of volunteering for a high-risk mission when others in their service academy cohort asks them to join the mission and fight together. Standardized Likert scale from -3 (very unlikely) to 3 (very likely).

* obeyorder: The willingness to obey the superior’s command even when one thinks it is unreasonable. Measured as the perceived appropriateness of the hypothetical command to fight a battle when a respondent’s platoon has already fought five high-risk missions. Standardized Likert scale from -3 (very inappropriate) to 3 (very appropriate).

** First set of DV scenario variation questions

* objective: The extent to which a respondent's answer to the first DV measuring task they completed is likely to change when the national objective of war changes. Likert scale from 1 (none at all) to 5 (a great deal).

* enemy: The extent to which a respondent's answer to the first DV measuring task they completed is likely to change when the enemy at war changes. Likert scale from 1 (none at all) to 5 (a great deal).

* risk: The extent to which a respondent's answer to the first DV measuring task they completed is likely to change when the accompanied risk of war changes. Likert scale from 1 (none at all) to 5 (a great deal).

* tie: The extent to which a respondent's answer to the first DV measuring task they completed is likely to change when the degree of social tie with others within their unit changes. Likert scale from 1 (none at all) to 5 (a great deal).

** First set of DV measurement validity questions

* dims_1: The extent to which a respondent's answer to the first DV measuring task they completed represents their order compliance. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

* dims_2: The extent to which a respondent's answer to the first DV measuring task they completed represents their willingness to use discretion in combat. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

* dims_3: The extent to which a respondent's answer to the first DV measuring task they completed represents their willingness to accept reasonable risk to their own life in combat. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

* dims_4: The extent to which a respondent's answer to the first DV measuring task they completed represents their willingness to accept reasonable risk to their subordinates' lives in combat. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

* dims_5: The extent to which a respondent's answer to the first DV measuring task they completed represents their willingness to cooperate with other units. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

* dims_6: The extent to which a respondent's answer to the first DV measuring task they completed represents their unit cohesion. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

** Second set of DV measuring tasks

* riskownlife2: The extent to which a respondent is willing/reluctant to risk their own life by volunteering to serve in a risky military mission. Standardized Likert scale ranging from -3 (strongly reluctant to) to 3 (strongly willing to).

* risktheirlife2: The extent to which a respondent is willing/reluctant to risk their subordinates’ lives by ordering them to serve in a risky military mission. Standardized Likert scale ranging from -3 (strongly reluctant to) to 3 (strongly willing to).

* discretion2: The extent to which a respondent is willing/reluctant to use their discretion to accomplish the operational objectives. Standardized Likert scale ranging from -3 (strongly reluctant to) to 3 (strongly willing to).

* coordination2: The extent to which a respondent trusts/mistrusts other units and is willing/reluctant to cooperate with them to accomplish the operational objectives by doing what they promised to do (advancing at 11pm in the given scenario). Standardized Likert scale ranging from -3 (I would almost definitely send a lookout and advance if others are.) to 3 (I would almost definitely advance at 11pm.).

* unitcohesion2: Unit cohesion, measured by the likelihood of volunteering for a high-risk mission when others in their service academy cohort asks them to join the mission and fight together. Standardized Likert scale from -3 (very unlikely) to 3 (very likely).

* obeyorder2: The willingness to obey the superior’s command even when one thinks it is unreasonable. Measured as the perceived appropriateness of the hypothetical command to fight a battle when a respondent’s platoon has already fought five high-risk missions. Standardized Likert scale from -3 (very inappropriate) to 3 (very appropriate).

** Second set of DV scenario variation questions

* objective2: The extent to which a respondent's answer to the second DV measuring task they completed is likely to change when the national objective of war changes. Likert scale from 1 (none at all) to 5 (a great deal).

* enemy2: The extent to which a respondent's answer to the second DV measuring task they completed is likely to change when the enemy at war changes. Likert scale from 1 (none at all) to 5 (a great deal).

* risk2: The extent to which a respondent's answer to the second DV measuring task they completed is likely to change when the accompanied risk of war changes. Likert scale from 1 (none at all) to 5 (a great deal).

* tie2: The extent to which a respondent's answer to the second DV measuring task they completed is likely to change when the degree of social tie with others within their unit changes. Likert scale from 1 (none at all) to 5 (a great deal).

** Second set of DV measurement validity questions

* dims2_1: The extent to which a respondent's answer to the second DV measuring task they completed represents their order compliance. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

* dims2_2: The extent to which a respondent's answer to the second DV measuring task they completed represents their willingness to use discretion in combat. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

* dims2_3: The extent to which a respondent's answer to the second DV measuring task they completed represents their willingness to accept reasonable risk to their own life in combat. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

* dims2_4: The extent to which a respondent's answer to the second DV measuring task they completed represents their willingness to accept reasonable risk to their subordinates' lives in combat. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

* dims2_5: The extent to which a respondent's answer to the second DV measuring task they completed represents their willingness to cooperate with other units. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

* dims2_6: The extent to which a respondent's answer to the second DV measuring task they completed represents their unit cohesion. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

** Seriousness of treatment scenarios

* seriousness_1:  The extent to which a respondent would take seriously a military training simulation based on the Chinese invasion of Taiwan and the US call for allied military support scenario. Standardized Likert scale from -2 (not at all seriously) to 2 (very seriously).

* seriousness_2: The extent to which a respondent would take seriously a military training simulation based on the war supported by the liberal party and not by the conservative party scenario. Standardized Likert scale from -2 (not at all seriously) to 2 (very seriously).

* seriousness_3: The extent to which a respondent would take seriously a military training simulation based on the Chinese occupation of the North Korean territory scenario. Standardized Likert scale from -2 (not at all seriously) to 2 (very seriously).

* seriousness_4: The extent to which a respondent would take seriously a military training simulation based on the battle with tamed dragons scenario. Standardized Likert scale from -2 (not at all seriously) to 2 (very seriously).

-------

*** measures_pilot_actual: Second round of pilot survey with Korean Army cadets. Used for checking the plausibility/seriousness of treatment scenarios and the measurement validity of each question intended to measure each of six dimensions of military cohesion.

** DV measurement validity questions

* measure_risklife_1: The extent to which respondents think a soldier's response to the "risklife" question would vary depending on their degree of order compliance. Standardized Likert scale from -2 (none at all) to 2 (a great deal). 

* measure_risklife_2: The extent to which respondents think a soldier's response to the "risklife" question would vary depending on their degree of willingness to exercise discretion. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

* measure_risklife_3: The extent to which respondents think a soldier's response to the "risklife" question would vary depending on their degree of willingness to risk their life in a battlefield mission. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

* measure_risklife_4: The extent to which respondents think a soldier's response to the "risklife" question would vary depending on their degree of willingness to risk their subordinates' lives in a battlefield mission. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

* measure_risklife_5: The extent to which respondents think a soldier's response to the "risklife" question would vary depending on their degree of willingness to coordinate with other units. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

* measure_risklife_6: The extent to which respondents think a soldier's response to the "risklife" question would vary depending on their degree of unit cohesion. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

* measure_risksubordin_1: The extent to which respondents think a soldier's response to the "risksubordin" question would vary depending on their degree of order compliance. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

* measure_risksubordin_2: The extent to which respondents think a soldier's response to the "risksubordin" question would vary depending on their degree of willingness to exercise discretion. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

* measure_risksubordin_3: The extent to which respondents think a soldier's response to the "risksubordin" question would vary depending on their degree of willingness to risk their life in a battlefield mission. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

* measure_risksubordin_4: The extent to which respondents think a soldier's response to the "risksubordin" question would vary depending on their degree of willingness to risk their subordinates' lives in a battlefield mission. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

* measure_risksubordin_5: The extent to which respondents think a soldier's response to the "risksubordin" question would vary depending on their degree of willingness to coordinate with other units. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

* measure_risksubordin_6: The extent to which respondents think a soldier's response to the "risksubordin" question would vary depending on their degree of unit cohesion. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

* measure_discretion_1: The extent to which respondents think a soldier's response to the "discretion" question would vary depending on their degree of order compliance. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

* measure_discretion_2: The extent to which respondents think a soldier's response to the "discretion" question would vary depending on their degree of willingness to exercise discretion. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

* measure_discretion_3: The extent to which respondents think a soldier's response to the "discretion" question would vary depending on their degree of willingness to risk their life in a battlefield mission. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

* measure_discretion_4: The extent to which respondents think a soldier's response to the "discretion" question would vary depending on their degree of willingness to risk their subordinates' lives in a battlefield mission. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

* measure_discretion_5: The extent to which respondents think a soldier's response to the "discretion" question would vary depending on their degree of willingness to coordinate with other units. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

* measure_discretion_6: The extent to which respondents think a soldier's response to the "discretion" question would vary depending on their degree of unit cohesion. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

* measure_coodrinate_1: The extent to which respondents think a soldier's response to the "interunitcoordination" question would vary depending on their degree of order compliance. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

* measure_coodrinate_2: The extent to which respondents think a soldier's response to the "interunitcoordination" question would vary depending on their degree of willingness to exercise discretion. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

* measure_coodrinate_3: The extent to which respondents think a soldier's response to the "interunitcoordination" question would vary depending on their degree of willingness to risk their life in a battlefield mission. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

* measure_coodrinate_4: The extent to which respondents think a soldier's response to the "interunitcoordination" question would vary depending on their degree of willingness to risk their subordinates' lives in a battlefield mission. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

* measure_coodrinate_5: The extent to which respondents think a soldier's response to the "interunitcoordination" question would vary depending on their degree of willingness to coordinate with other units. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

* measure_coodrinate_6: The extent to which respondents think a soldier's response to the "interunitcoordination" question would vary depending on their degree of unit cohesion. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

* measure_unitco_1: The extent to which respondents think a soldier's response to the "unitcohesion" question would vary depending on their degree of order compliance. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

* measure_unitco_2: The extent to which respondents think a soldier's response to the "unitcohesion" question would vary depending on their degree of willingness to exercise discretion. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

* measure_unitco_3: The extent to which respondents think a soldier's response to the "unitcohesion" question would vary depending on their degree of willingness to risk their life in a battlefield mission. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

* measure_unitco_4: The extent to which respondents think a soldier's response to the "unitcohesion" question would vary depending on their degree of willingness to risk their subordinates' lives in a battlefield mission. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

* measure_unitco_5: The extent to which respondents think a soldier's response to the "unitcohesion" question would vary depending on their degree of willingness to coordinate with other units. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

* measure_unitco_6: The extent to which respondents think a soldier's response to the "unitcohesion" question would vary depending on their degree of unit cohesion. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

* measure_obey_1: The extent to which respondents think a soldier's response to the "obeyorder" question would vary depending on their degree of order compliance. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

* measure_obey_2: The extent to which respondents think a soldier's response to the "obeyorder" question would vary depending on their degree of willingness to exercise discretion. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

* measure_obey_3: The extent to which respondents think a soldier's response to the "obeyorder" question would vary depending on their degree of willingness to risk their life in a battlefield mission. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

* measure_obey_4: The extent to which respondents think a soldier's response to the "obeyorder" question would vary depending on their degree of willingness to risk their subordinates' lives in a battlefield mission. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

* measure_obey_5: The extent to which respondents think a soldier's response to the "obeyorder" question would vary depending on their degree of willingness to coordinate with other units. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

* measure_obey_6: The extent to which respondents think a soldier's response to the "obeyorder" question would vary depending on their degree of unit cohesion. Standardized Likert scale from -2 (none at all) to 2 (a great deal).

** Seriousness of treatment scenarios

* seriousness_1: The extent to which a respondent would take seriously a military training simulation based on the Chinese invasion of Taiwan and the US call for allied military support scenario. Standardized Likert scale from -2 (not at all seriously) to 2 (very seriously).

* seriousness_2: The extent to which a respondent would take seriously a military training simulation based on the war supported by the liberal party and not by the conservative party scenario. Standardized Likert scale from -2 (not at all seriously) to 2 (very seriously).

* seriousness_3: The extent to which a respondent would take seriously a military training simulation based on the Chinese occupation of the North Korean territory scenario. Standardized Likert scale from -2 (not at all seriously) to 2 (very seriously).

-------

*** me: Expert survey of the construct validity of military effectiveness (later interpreted as military cohesion, as we sharpened our use of the concept). Conducted with senior Korean commissioned Army officers.

** Military effectiveness indicators

* me1: The extent to which a respondent believes that order compliance is an important factor of military effectiveness. Standardized Likert scale from -3 (not important at all) to 3 (Very important).

* me2: The extent to which a respondent believes that the use of discretion on battlefield to accomplish operational objectives is an important factor of military effectiveness. Standardized Likert scale from -3 (not important at all) to 3 (Very important).

* me3: The extent to which a respondent believes that physical fitness is an important factor of military effectiveness. Standardized Likert scale from -3 (not important at all) to 3 (Very important).

* me4: The extent to which a respondent believes that shooting skill is an important factor of military effectiveness. Standardized Likert scale from -3 (not important at all) to 3 (Very important).

* me5: The extent to which a respondent believes that the willingness to risk their own life in combat to accomplish operational objectives is an important factor of military effectiveness. Standardized Likert scale from -3 (not important at all) to 3 (Very important).

* me6: The extent to which a respondent believes that the willingness to risk their subordinates' lives in combat to accomplish operational objectives is an important factor of military effectiveness. Standardized Likert scale from -3 (not important at all) to 3 (Very important).

* me7: The extent to which a respondent believes that the capability to coordinate and cooperate with other units in an operation is an important factor of military effectiveness. Standardized Likert scale from -3 (not important at all) to 3 (Very important).

* me8: The extent to which a respondent believes that the trust in other members of their unit is an important factor of military effectiveness. Standardized Likert scale from -3 (not important at all) to 3 (Very important).

* open: Responses to an open-ended question asking what other factors not mentioned as a military effectiveness indicator in the survey would constitute military effectiveness.

** Demographics

* age: Whether a respondent is in their thirties (0), forties (1), or fifties (2).

* rank: Whether a respondent's current military rank is major (0), lieutenant colonel (1), colonel (2), or general (3).

* mil_yr: The number of years a respondent has served in the military.

* field_yr: The number of years a respondent has been stationed in the military base as a field officer.

* gender: 0 if a respondent’s biological sex given by birth is female and 1 if male. 999 if prefer not to state.

* hometown:  Identifier for the province in which a respondent’s hometown is located. We intend to capture the regional divide between Gyeongsang province and Jeolla province, which are known for the long history of hostility against each other. 2 if a respondent is from Jeolla province, 1 if from Gyeongsang province, and 0 if from the rest (including foreign countries).

* militant_assertiveness: Hawkishness measured by the extent to which a respondent agrees/disagrees that the use of force only makes problems worse. Standardized Likert scale from -3 (strongly agree) to 3 (strongly disagree).
